【作者】 陈福年；

【导师】 钱真；

【作者基本信息】 哈尔滨工程大学 ， 计算机应用技术， 2011， 硕士

【摘要】 随着信息时代的到来,电子地图被广泛应用于各个领域。电子地图给我们带来方便的同时也带来了许多问题,比如地图的非法传播,地图的版权保护。数字水印作为一种重要的版权保护方法,已经被广泛的应用于多种数字产品的版权保护。本文提出了一种适用于影像地图的水印方法,该算法以二值位图图像作为水印信息,以地图作为水印载体。本文主要研究了水印技术在影像地图的版权保护中的应用,提出了基于小波域隐马尔科夫树的盲水印算法。该算法首先利用影像地图的地理属性对影像地图进行网格划分,选取重要的地图分块进行小波变换。然后根据小波系数的特点,利用隐马尔科夫模型对小波系数进行分析,提取重要的小波系数。最后采用本文提出的步长量化算法对小波系数进行扰动。本文通过分析了现有模型应用于水印算法的不足之处,提出了对隐马尔科夫模型的改进方法。通过在原步长量化算法的基础上加入了容错机制,对步长量化算法进行了较大改进。本文提出的水印算法有效的利用了影像地图的特点,对原有算法进行了改进。仿真实验结果表明,本算法具有较好的透明性,较高的鲁棒性,能有效抵抗噪声、压缩、剪切等攻击。更多还原

【Abstract】 With the advent of the information age, digital map has been widely used in various fields. Digital map has brought us convenience but also there are many problems need to be solved. For example, illegal distribution of maps, map copyright protection. As an important method for copyright protection, digital watermarking has been widely used in copyright protection for many digital products. A kind of watermarking algorithm is proposed for remote sensing image. The binary bitmap is used as the watermark information, and the map is used as the watermark carrier.We mainly researched watermark technology application for map copyright protection in this paper. We proposed a blind watermarking algorithm for Remote Sensing Image based on wavelet domain and Hidden Markov Tree in this paper. Firstly, this algorithm makes use of the geographical attributes of the map to cut the map into small blocks. Then, we selected the important blocks and uses wavelet transform on the blocks. After that, we make use of Hidden Markov Model to analysis the wavelet coefficients and extract important coefficients according to the characteristics of wavelet coefficients. Finally, we make use of the proposed quantization algorithm to modify wavelet coefficients. We proposed improvement of the Hidden Markov Model by analyses of the inadequacies of existing models. We improved the quantization algorithm by adding the fault tolerance based on the original quantization algorithm.The algorithm, which was improved based on the Original algorithm, makes good use of the characteristics of the remote sensing image. Simulation experiments results show that this algorithm has good transparency and strong robustness. It can effectively resist attacks like noise, compression, clipping and so on.更多还原